Hacker News

New stories

Live mirror
30 storiesupdated just nowView source snapshot
  1. U.S. Rebukes Australia over Proposed Limits on Social Media Algorithms (nytimes.com)
    —discuss
  2. Microsoft Abandons Personal AI Chatbot Race with Copilot Reboot (bloomberg.com)
    —discuss
  3. Show HN: Vitals – A Mac activity monitor that thinks in apps, not processes (vitalsmac.com)
    —discuss
  4. Allow Carriers on Planes (jefftk.com)
    —discuss
  5. Lost Egypt Volumes 1–3 (19th CE) (uchicago.edu)
    —discuss
  6. Not Everyone Can Code (funcall.blogspot.com)
    —discuss
  7. ReAnchor, a Jev Optimizer in DSPy (cmpnd.ai)
    —discuss
  8. A Jevlike using an escalation model to route between a classifier and an LLM (yogthos.net)
    —discuss
  9. SpaceX-Focused Pentagon Contracts Leave Rivals Feeling Squeezed (bloomberg.com)
    1comments
  10. First Principles Thinking (sunilsadasivan.com)
    —discuss
  11. The US stood alone in dismissing AI safety concerns at the UN (semafor.com)
    1comments
  12. Benchmarking Jev, Laya, and five open models under three stresses (github.com/gazelle93)
    —discuss
  13. ASML says it sold 'absolutely nothing' in Europe in 2026 (tomshardware.com)
    3comments
  14. Does Georgism Work? Five Years Later (astralcodexten.com)
    —discuss
  15. 30 Years of Backwards Compatibility in a Development Tool Since 1993 (visualneo.com)
    —discuss
  16. Tell agents the why, not just the how (seangoedecke.com)
    —discuss
  17. What happens when you analyze your favorite college football team like the CIA? (cultivatelabs.com)
    1comments
  18. Cracks in Oracle's AI Data-Center Build-Out Appear in New Mexico Project (wsj.com)
    —discuss
  19. CommonGrid: The open source registry of the US power grid (texturehq.com)
    —discuss
  20. The state of SIMD in Rust in 2026 (shnatsel.github.io)
    —discuss
  21. XDude (xdude.com)
    —discuss
  22. My Take on Jev (sunbeams.ai)
    —discuss
  23. Evaluating the Tax Character of Options-Based Income [pdf] (sterlingcapital.com)
    —discuss
  24. A Model Context Protocol (MCP) server for NATS messaging system integration (github.com/sinadarbouy)
    —discuss
  25. The New Copilot with Home, Code and Autopilot (blogs.microsoft.com)
    —discuss
  26. Show HN: Sifthound – self-hosted, Tavily-compatible search API for AI agents (github.com/khsarvar)
    —discuss
  27. You need to pass all unhandled messages to DefWindowProc, part 2 (devblogs.microsoft.com/oldnewthing)
    —discuss
  28. When working with Win32 FILETIME, don't forget that you have std:chrono now (devblogs.microsoft.com/oldnewthing)
    —discuss
  29. Feast feature and vector serving at ScyllaDB speed (feast.dev)
    1comments
  30. What's the highest legal FILETIME? Is it safe to use? (devblogs.microsoft.com/oldnewthing)
    —discuss

Anthropic: The Situation Report

25 pointsby 3h agoanthropic.com
11 comments
2h agoHN ↗

In before the inevitable, endless cynicism that always floods this kind of thread. This seems like a good thing, and I enjoyed reading about it.

2h agoHN ↗

I don't care if this is a PR stunt or a highly exxaggerated contribution of claude in the entire ebola tracking scenarios, I feel very optimistic about the future around how these models are just becoming a daily life tool for everyone and not just programmers or tech bros

2h agoHN ↗

Tldr: Claude makes reports and dashboards for ebola health workers

2h agoHN ↗

I think they’d rather be problem-solving messy human logistics than putting together presentations!

2h agoHN ↗

Claude is helping the organization (people) react faster, and more accurately.

The article mentions: you defeat Ebola by meeting it where it is now, not where it was. You can only beat it by stopping the spread of the infection, so it's very critical to move with speed and accuracy. Dashboards and reporting aren't very glamorous, I know, but we're really dealing with the flow of information to stop the spread of infection.

Intelligence is the ability to predict, and we need information to do it. It wasn't directly said in the article, but the information allows you to see patterns and predict spread, and meet Ebola 'now'.

As in, imagine the moment in time where you predict Ebola will be, and then take every measure to have an advantage over it. The reports and dashboards help you to imagine, predict that future 'now' moment.

If you don't have accurate information then you are always behind the disease chasing its spread.

1h agoHN ↗

it's very critical to move with speed and accuracy

Traditionally one of the weak points with this tech has been accuracy, are there new techniques to increase or verify the validity of the dashboards being generated?

2m agoHN ↗

You are correct, and I don't know. The only way to know is to listen to the people using the tool, but also measure the results. So you would need to compare past, similar responses to today's response, and factor in any new techniques/knowledge used in fighting the disease to accurately compare.

1h agoHN ↗

Why are we there when the locals hate the aid workers and actively try to fight them?

Because public health is about protecting everyone. You sound like you don't really care about the philosophical altruistic view, but the cynical reasoning works just as well here: if you can contain it in-situ, in Africa, it stops it spreading to the countries paying for the aid. Instead of gratitude from the direct recipients for how your taxes are being spent, you can instead bask in the warm glow of not catching ebola yourself.

2h agoHN ↗

Probably one of the better real-world AI use cases I’ve seen. Automate the tedious data work, save hours, and keep the actual scientific decisions with experts.

12m agoHN ↗

More descriptive title maybe: How health organizations are using Claude to respond to Ebola spread in the DRC